Customer Retention Command

Inspiration

Customer churn is one of the biggest revenue leaks for businesses. We wanted to create a command center that doesn’t just predict churn, but also takes real-time action to retain customers before they are lost.

What it does

Customer Retention Command is an AI-powered command center that helps businesses reduce churn and protect revenue.

Detects High-Risk Customers: Uses churn features and ML predictions to flag customers most likely to leave.
Estimates At-Risk Revenue: Calculates potential revenue loss tied to high-risk customers.
Tracks Average Risk Score: Provides a clear KPI to monitor overall customer retention health.
Visualizes Insights in Tableau: Interactive dashboards show high-risk segments by age, city, gender, and revenue distribution by offer tiers.
Automates Retention Workflows: Connects Tableau insights with Salesforce Flow and Slack to automatically post alerts and create follow-up tasks for sales/service teams.
Agentic Workflow: Uses Prompt Builder and Salesforce Data Cloud to power AI-driven, real-time recommendations.

In short, it transforms raw customer data into actionable insights + automated workflows, giving teams the tools they need to act faster, reduce churn, and retain more customers.

How we built it

Data Cloud → Loaded CSV into a data stream & semantic model.
Einstein Discovery Fast Track → Generated churn prediction model.
Tableau KPIs & Visuals →
- KPI-High-Risk Customers
- KPI-At-Risk Revenue
- KPI-Average Risk Score
- Bar Chart → High-Risk by Segment
- Donut Chart → At-Risk Revenue by Offer
Agentic Workflow → Prompt Builder + Flow + Slack.
Example: If High_Risk_Flag = 1 → Post to Slack & Create Salesforce Task.
[ Risk_Score > 0.5 \; \Rightarrow \; Action = "Retention Workflow" ] ## Challenges we ran into
Resolving duplicate API names when creating metrics.
Aligning semantic model fields with calculated KPIs.
Building a clean dashboard layout under time pressure.